摘要
为了均衡利用容器云集群的资源,对Mesos的资源调度策略做了改进。改进的资源调度策略是在原有调度策略的基础上,把Mesos-Slave节点的CPU使用率、内存的占有率以及运行的容器个数考虑进去,并对三者加权来做为接下来资源调度的依据。最后通过改进前后集群资源利用率的对比,验证了改进后资源调度策略能够使集群资源利用率趋于均衡。
In order to balance the use of container cloud cluster resources,the resource scheduling strategy of Mesos made improvements. Improved resource scheduling policy is based on the original scheduling policy and the CPU usage,share memory and the number of containers of Mesos-Slave nodes run into account,and to do for the next three weighted resource scheduling in accordance with. Finally,by comparing before and after the cluster improved resource utilization,validate the improved resource scheduling strategy can make the cluster resource utilization tends to be balanced.
出处
《计算机与数字工程》
2017年第10期1931-1936,共6页
Computer & Digital Engineering
关键词
容器云
集群资源利用率
资源调度策略
container cloud
cluster resource utilization
resource management strategy
